Objective: To determine the prevalence of gestational diabetes mellitus in a large general obstetric population and its variations depending on the presence of risk factors, and to evaluate how the gestational diabetes screening strategies applied might modify the observed prevalence in the population. Design: A retrospective cohort study. Methods: The study population was a total of 2574 pregnant women. Information about risk factors, screening and diagnosis of gestational diabetes was obtained. Frequency of risk factors under the American College of Obstetrics and Gynecologists (ACOG) and the American Diabetes Association (ADA) criteria, and observed and expected prevalence of gestational diabetes mellitus were calculated and compared for statistical significance. Results: Age $ 30 years, family history of diabetes, obesity and previous fetal macrosomia were the most frequent risk factors. Under ACOG recommendations, 45% of our general obstetric population would have been exempt from gestational diabetes mellitus screening, as compared with only 15.5% under ADA guidelines. Sixty-five patients were diagnosed as having gestational diabetes mellitus, giving an overall prevalence of 2.5% (confidence interval 2.0 -3.2). Among the low-risk women, prevalence values were 0.6% and 0.5% respectively under ACOG and ADA criteria, whereas for those presenting one or more risk factors rates were 4% and 2.9% respectively. Conclusions: In our general obstetric population, gestational diabetes mellitus prevalence was found to be approximately six times lower among low-risk gravidae than among the high-risk subjects, suggesting that selective screening might be beneficial. Nevertheless, selective gestational diabetes mellitus screening under ADA criteria seems to entail the same disadvantages as the selective screening strategies without any apparent benefits.